Skip to main content

Full text of "Pandion lovensis new species of Ospreay Florida deposits"

See other formats


\jP Biodiversity 
fe^HeriUge 

http://www.biodiversitylibrary.org 



Proceedings of the Biological Society of Washington. 

Washington, Biological Society of Washington 
http://www.biodiversitylibrary.org/bibliography/3622 



V. 98 1985: http://www.biodiversitylibrary.org/item/107750 

Page(s): Page 314, Page 31 5, Page 316, Page 31 7, Page 31 8, Page 31 9, Page 320, Page 544, 

Page 545, Page 546, Page 547, Page 548, Page 549, Page 550, Page 551 , Page 552, Page 

553 



Contributed by: Smithsonian Institution Libraries 
Sponsored by: Biodiversity Heritage Library 



Generated 24 May 201 1 4:48 PM 
http://www.biodiversitylibrary.org/pdf3/006803200107750 



This page intentionally left blank. 




PROC. BIOL. SOC. WASH. 
98(2), 1985, pp. 314-320 



PANDION LOVENSIS, A NEW SPECIES OF OSPREY FROM 

THE LATE MIOCENE OF FLORIDA 



Jonathan J. Becker 



Abstract.— Pandion lovensis n. sp. (Aves: Pandionidae) is described from the 
late Miocene (latest Clarendonian) of Florida. This species, based on pelvic limb 
elements, appears to be the most primitive member of the genus. 



The modem osprey {Pandion haliaetus) is the sole living representative of an 
enigmatic family of diurnal raptors. A number of detailed studies have investigated 
the morphology of the osprey in order to clarify its taxonomic position (Compton 
1938; Hudson 1948; Sibley and Ahlquist 1972; Jollie 1976-1977). These studies 
have placed the osprey in a separate family or suborder, usually allied with the 
hawks and eagles. 

Warter (1976) discussed the fossil record of the Pandionidae and described the 
first paleospecies, Pandion homalopteron, from Sharkstooth Hill, but did not find 
any convincing evidence to ally the modem osprey closely with any other fal- 
coniform group. Sharkstooth Hill near Bakersfield, Kern County, California, is 
early middle Miocene (about 14.5-13 million years B.P.) and is "closely tied into 
the late 'Temblor' megain vertebrate stage and the Luisian microin vertebrate stage" 
(Repenning and Tedford 1977:79). Pandion homalopteron, based on associated 
humeri and ulnae, represents an osprey slightly larger than the average modem 
osprey, that shows evidence of weaker wing musculature (Warter 1976). 

Brunet (1 970) proposed the transfer of Palaeocircus cuvieri Milne-Edwards from 
the Accipitridae to the Pandionidae, but because of the incompleteness of the 
holotype, a fragmentary carpometacarpus, this has not been accepted by other 
workers (Warter 1976). 

Reported here, from the late Miocene of Florida is the second known paleo- 
species of osprey. 

Abbreviations. —Sptcim^ns cited below are housed in the following institutions: 
Florida State Museum (UF), collection of Pierce Brodkorb (PB), and Natural 
History Museum of Los Angeles County (LACM). 

Recent specimens examined.— Pandion haliaetus carolinensis four male speci- 
mens, PB 20312, PB 39212, PB 27958, UF 19406; four female specimens, PB 
17061, UF 14546, UF 17082, UF 18215; four specimens of unknown sex, PB 
34670, PB 39613, PB 37976, PB 34669. 

Fossil specimens examined.— Kcferr^d proximal end of left tibiotarsus of Pan- 
dion homalopteron (LACM 42815). Subsequent to Warter's (1976) description of 
this species, this tibiotarsus was collected by Mr. William Hawes from the same 
location at Sharkstooth Hill (LACM locality 3205) that he collected the type- 
material of P. homalopteron (L. G. Barnes, in litt. 1982). It should be noted, that 
although this specimen bears the same catalog number as the holotype, it cannot 
be considered type-material (except as a referred hypotype), as it was not included 
in the original description. 



VOLUME 98, NUMBER 2 



315 



Descriptive statistics are based on all above recent specimens. Morphological 
comparisons are based on the seven specimens in the Brodkorb collection. Mea- 
surements (Table 1) were made with Kanon dial calipers, accurate to 0.05 mm 
and rounded to the nearest 0. 1 mm. BMDP Statistical Software program BMDPID 
was used to calculate simple descriptive statistics (Dixon 1981). Computations 
were made at the Northeast Regional Data Center (NERDC) at the University of 
Florida, Gainesville. All fossil specimens are deposited in the Vertebrate Paleon- 
tology collections of the Rorida State Museum, University of Florida (UF). An- 
atomical terminology follows Baumel et al. (1979) and Howard (1929). 



Order Accipitriformes (Falconiformes auct.) 
Family Pandionidae (Sclater and Salvin, 1893) 

Skeletal elements of pelvic limb distinguished from other accipitriform families 
by the following combination of characters: (1) femur with very deep popliteal 
fossa; (2) tibiotarsus with extensor canal very deep under tendinal bridge with 
single distal opening; (3) fibula fused far distad; (4) tarsometatarsus relatively 
short, with ossified retinaculi extensoris for M. extensor digitorum longus; (5) 
hypotarsus extremely large with a single circular opening for tendons of Mm, 
flexor digitorum longus and flexor hallucis longus; (6) calcaneal ridge grooved; (7) 
trochleae strongly arched; (8) claws rounded beneath. 



Genus PandionSavigny , 1809 
Pandion lovensis, new species 

Figs. 1, 2 

Holotype.— Nearly complete left tarsometatarsus. Vertebrate Paleontology col- 
lections of the Horida State Museum, UF 25950 (Fig. lb, c); collected in 1979 
by personnel of the Florida State Museum. 

Type- Locality.— 1.0YC Bone Bed local fauna. Florida, Alachua County, along 
State Road 24 1 , NW Va, SW Va, NW Va, Sec. 9, T. 1 1 S., R. 1 8 E., Archer Quadrangle, 
U.S. Geological Survey 7.5 minute series topographical map, 1969. Webb et al. 
(1981) give an overview of this local fauna. 

Horizon. —Late Miocene, latest Clarendonian land mammal age (approximately 
9 million years B.P.). The Love Bone Bed local fauna originates from fluvial 
sediments of the Alachua Formation (Williams et al. 1977). 

Etymology.— For the type locality, the Love Bone Bed. 

Paratypes.—jyistal half of right femur, UF 25766; distal end of right tibiotarsus, 
UF 25884; complete left tibiotarsus, UF 25928; right tarsometatarsus lacking 
proximal end, UF 25863; ungual phalanges, UF 26055, UF 26056, UF 29660. 

Measurements.— Table 1. 

Distinguished from P. haliaetus by; longer and more slender tar- 
sometatarsus, lateral proximal vascular foramen opening within hypotarsal canal; 
femur with patellar sulcus broader and caudal intermuscular line more mediad; 
tibiotarsus with anterior and posterior intercondylar sulci wider and less deep, 
cranial opening of extensor canal larger and more transversely oriented, and distal 
end wider. Distinguished from P. homalopteron by a tibiotarsus with smaller 
transverse width of proximal end and deeper fossa retrocristalis. 





316 



PROCEEDINGS OF THE BIOLOGICAL SOCIETY OF WASHINGTON 



Table 1 .— Measurements of Pandion spp. Data are number of specimens (n), mean ± standard 
deviation (Jc ± SD) and range. Tibiotarsus— A, Total length; B, Length fibular crest; C, Least depth 
shaft; D, Depth proximal end; E, Transverse width proximal end; F, Transverse width distal end, 
across anterior portion of condyles; G, Transverse width distal end, across posterior portion of condyles; 
H, Depth medial condyle; I, Depth lateral condyle; J, Least depth intercondylar area. Tarsometatar- 
sus— K, Total length, from eminentia intercondylaris through trochlea III; L, Length metatarsal I 
facet; M, Transverse width trochlea III; N, Transverse width distal end; O, Depth trochlea III; P, 
Transverse width proximal end; Q, Depth proximal end, excluding hypotarsus. Femur— R, Transverse 
width of lateral condyle; S, Transverse width of medial and lateral condyles; T, Transverse width 
lateral condyle and trochlea fibularis; U, Transverse width distal end; V, Depth distal end; W, Depth 
femoral shaft cranial to condyles. 



Measure- 
ment 



P. h. carolinensis 



jc ± SD 



(n) 



Range 



P. 

homalopteron 



p. lovensis 



Tibiotarsus 

A 
B 

C 

D 

E 

F 

G 

H 

I 

J 



123.59 ± 4.62 
34.82 ± 2.09 



5.59 ± 0.29 



17.00 ± 0.92 



13.17 ± 0.60 
14.08 ± 0.69 
10.36 ± 0.50 

13.18 ± 0.68 
12.77 ± 0.57 

5.83 ± 0.29 



Tarsometatarsus: 



K 
L 

M 
N 
O 
P 

Q 

Femur: 

R 

S 
T 
U 

V 

w 



51.86 



6.81 



13.51 



1.64 



9.38 ± 0.80 

6.80 ± 0.56 

15.00 ± 0.54 



4.98 ± 0.26 

14.46 ± 0.76 

5.49 ± 0.28 



3.12 ± 0.27 



12.14 ± 0.69 



0.56 



15.29 ± 0.89 



0.47 



7.70 ± 0.35 



(12) 

(13) 
(12) 

(13) 
(13) 
(12) 
(12) 
(12) 
(12) 
(12) 




(12) 
(12) 
(12) 
(12) 
(12) 




(13) 

(13) 
(13) 

(13) 
(13) 
(13) 



119.2-130.8 
31.2-38.2 

5.2-6.1 
15.9-18.5 
12.3-14.0 
13.1-15.1 

9.6-11.1 
12.4-14.2 
12.0-13.6 

5.4-6.5 



49.7-54.9 
8.0-10.6 
5.8-7.6 

14.4-15.9 
4.6-5.5 

13.4-15.7 

5.1-6.0 



2.8-3.6 
11.2-12.9 

6.0-7.9 
14.0-16.6 
12.7-14.2 

7.2-8.3 



17.4 
14.3 



124.8 

32.7 

5.5; 5.7 
17.0 
13.1 

14.9; 15.0 
12.0; 12.1 
13.2; 13.3 
12.0; 12.5 

6.6; 6.6 



59.5 
8.0; 8.4 

7. 8; 7.0 
16.4 

5.7; 5.7 
14.7 

6.7 



3.1 

12.4 

6.6 
15.2 
12.8 

7.2 



Comparisons and description.— Unless otherwise stated, all comparisons are 
made in relation to 7 specimens (Pierce Brodkorb collection) oi Pandion haliaetus 
carolinensis GuiQlin, 1788. 

Femur .— Pandion lovensis n. sp. has caudal intermuscular line more mediad, 
merging smoothly with the crista supracondylaris medialis, forming a sharp caudo- 
medial border immediately above the medial epicondyle. Caudal aspect of the 
medial condyle broader. Popliteal fossa slightly broader. Caudal aspect of lateral 
condyle (i.e., tibial articular surface) extending less craniad and is not inclined 
laterad. Crista tibiofibularis and lateral epicondyle less pronounced. Patellar sulcus 
slightly broader. 

Tibiotarsus,— Fibular crest shorter. Both anterior and posterior intercondylar 



VOLUME 98, NUMBER 2 



317 









\ 




I* ^ 



wfc ^^ 



.^; 



X. 






r 













A 




-V 



V 








Fig. 1. Stereophotographs of tarsometatarsi of Pandion lovensis n. sp. A, Paratype UF 25863, 
distal view. B, C. Holotype, UF 25950. B, Caudal view; C, Cranial view. Scale equals 10 mm. (A); 
20 mm. (B, C). 



sulci wider and less deep. Both lateral and medial epicondylar depressions deeper 
and more distinct. Cranial opening of extensor canal larger and more transversely 
oriented. Internal ligamental prominence more distinct. Distal end wider, espe- 
cially caudal portion. 

The referred proximal end of tibiotarsus of Pandion homalopteron (LACM 
42815), when compared with P. haliaetus, is robust, with a greater transverse 
width of proximal end. Crista cnemalis lateralis slightly elongated, producing a 
more pronounced incisura tibialis and a broader sulcus intercristalis. Facies gas- 





318 



PROCEEDINGS OF THE BIOLOGICAL SOCIETY OF WASHINGTON 




i 



i 






ft ^ 



'^ 








ii. ^'■*- 




V 








r 



^ 




k k 



■*/l 



\ 



A*-* 



Nr\. 



^ * t 



- 1 



Fig, 2. Paratypes of Pandion lovensis. A, Caudal view femur UF 25766; B, Cranial view tibiotarsus 
UF 25928; C, Caudal view tibiotarsus UF 25928. Scales = 10 mm. 



trocnemalis flatter. Distinct notch present on caudo -lateral margin of proximal 
articulating surface. Slight ridge extends proximad from tuberositas poplitea. 

In comparison with Pandion homalopteron, P. lovensis has a smaller transverse 
width, a less distinct notch on the caudo-lateral margin of the proximal end, and 
a distinctly deeper fossa retrocristalis. 

Tarsometatarus. —Shaft longer, more slender, and less flattened dorso-plantarly. 
Fossa parahypotarsalis medialis slightly more excavated. Crista medialis hypotarsi 
extends a proportionally shorter distance down shaft. Lateral foramen vascularia 
proximalia opens within hypotarsal canal (outside in all specimens of Pandion 
haliaetus examined). Fovea ligamentum coUateralis on trochlea IV larger and 
deeper. Trochlea IV less recurved, anterior surface flattened. Trochlea III larger. 
Medial foramen vascularia proximalia proximal to origin of inner strut of arcus 
extensoris. Fossa infracotylaris dorsalis deeper. Medial border of trochlea III 
projects laterad (dorsad in P. haliaetus). Distal end larger. In medial view, caudal 
process on trochlea II proportionally longer. 




VOLUME 98, NUMBER 2 



319 



Discussion. —The power-arm ratio of the tarsometatarsus has been the focus of 
many investigations (Miller 1911, 1912; Howard 1932; JoUie 1976-1977; among 
others). The major flexor of the tarsometatarsus on the shank is the M. tibialis 
anterior (=M. tibialis cranialis of Baumel et al. 1979) and, to a lesser degree, the 
M. extensor digitorum longus. In Pandion haliaetus, the tibialis anterior arises 
by two heads, a tibial head on the anterior side of the tibial crest, extending in a 
narrow line down the medial side of the tibial shaft; and a femoral head extending 
from the distal apex of the external condyle of the femur. This muscle inserts by 
a single tendon on the tibialis anterior tuberosity on the proximal end of the 
tarsometatarsus (Hudson 1937, 1948). 

The power-arm ratio (Miller 1912, 1925) is calculated by dividing the length 
from the proximal end to the midpoint of the tibialis anterior tubercle (=power- 
arm) multiplied by 100, by the total length of the tarsometatarsus (=resistance or 
weight-arm). Miller (cited in JoUie 1976-1977) noted that species with long tarsi 
have a short power-arm ratio while those species with a short broad tarsus have 
a relatively large ratio. Miller (1911), Howard (1932), and JoUie (1976-1977) 
provide tables of power-arm ratios for comparison. It is interesting to note that 
the modem osprey has the greatest power-arm ratio (32.2%) of any accipitriform 
species. Pandion lovensis has a much smaller power-arm ratio (17.0 mm/59.5 
mm X 100 = 28.6%). This is approximately 11% less than the modem osprey. 
An increase in length of the tarsometatarsus, without a concomitant shift in the 
position of the tibialis anterior tubercle is responsible for the decrease in the 
power-arm ratio in P. lovensis. The increase in length would also allow the distal 
end of the tarsometatarsus to be moved at a faster rate, all other things being 
equal. 

The interpretation of these differences is difficult. Fisher (1945:742) states "The 
development of this great flexor of the tarsus may be correlated with ability to 
walk or run, ability to grasp with the foot as in perching or in predation, and with 
weight of the foot or of the entire body. In fact it is impossible to define and 
distinguish individual adaptations." 

Because P. homalopteron, P. lovensis, and P. haliaetus have only one known 
skeletal element in common, any proposed phylogeny is tenuous. Pandion hom- 
alopteron is not very distinct in wing morphology from the modem osprey, even 
though a large interval of time separates them (Warter 1976). The only known 
hindlimb element of this species, a proximal end of a tibiotarsus, also appears 
close to that of the modem osprey. Pandion lovensis is less derived than either 
of these species and shares a number of characters with the Accipitridae, the 
proposed sister taxa of the Pandionidae (JoUie 1976-1977). These characters 
include a femur with a broader and less deep patellar sulcus, and the caudal 
intermuscular line medial; a tibiotarsus with broader intercondylar sulci; and a 
tarsometatarsus which is longer and less broad, with a reduced power-arm ratio. 

Pandion lovensis appears to be the least derived member of the genus and 
represents a lineage distinct from that of P. homalopteran and P. haliaetus. 



Acknowledgments 

I thank Pierce Brodkorb, Department of Zoology, University of Florida and 
the personnel of the Division of Ornithology, Florida State Museum for loan of 
skeletal specimens. S. David Webb and B. J. MacFadden, Division of Vertebrate 




320 



PROCEEDINGS OF THE BIOLOGICAL SOCIETY OF WASHINGTON 



Paleontology, Florida State Museum, University of Florida, and L. G. Barnes, 
Los Angeles County Museum of Natural History, made fossil specimens or casts 
available for study. I especially thank R. G. Wolff, Hildegarde Howard, Storrs L. 
Olson, Cecile Mourer-Chauvire and P. Brodkorb for their comments on the manu- 



script. 



Literature Cited 



Baumel, J. J., King, A. S., Lucas, A. M., Beazile, J. E., and Evans, H. E. (Eds.). 1979. Nomina 

Anatomica Avium. Academic Press, New York. 637 pp. 
Brunei, J. 1970. Oiseaux de I'eocene Superieur du Bassin de Paris.— Annates de Paleontologie 

(Vertebres) 56:1-57, plates A-D. 
Compton, L. V. 1938. The pterylosis of the Falconiformes with special attention to the taxonomic 

position of the Osprey.— University of California Publications in Zoology 42(3): 173-2 12. 
Dixon, W. J. 1981. BMDP Statistical Software. — University of California Press. Berkeley, California. 

725 pp. 
Fisher, H. I. 1945. Locomotion in the fossil vulture Teratornis.— American Midland Naturalist 

33(3):725-742. 
Howard, H. 1929. The Avifauna of Emeryville Shellmound. — University of California Pubhcations 

m Zoology 32:301-394. 
. 1932. Eagles and eagle-like vultures of the Pleistocene of Rancho La Brea. — Carnegie In- 



stitution of Washington, Publication 429:1-82, 29 pis. 
Hudson, G. E. 1937. Studies on the muscles of the pelvic appendage in birds.— American Midland 

Naturalist 18(1):1-108. 
. 1948. Studies on the muscles of the pelvic appendage in birds. IL The heterogenous order 



Falconiformes.— American Midland Naturalist 39(1): 102-1 27. 
JoUie, M. 1976-1977. A contribution to the morphology and phylogeny of the Falconiformes.— 

Evolutionary Theory 1:285-298, 2:115-300, 3:1-141. 
Miller, L. 1911. A series of eagle tarsi from the Pleistocene of Rancho La Brea.— University of 

California Publications, Bulletin of the Department of Geology 6(12):305-316. 
. 1912. Contributions to avian paleontology from the Pacific Coast of North America.— 



University of California Publications, Bulletin of the Department of Geology 7(5):61-1 15. 
. 1925. The birds of Rancho La Brea. — Carnegie Institution of Washington, Publication 349: 

63-106, 6 pis. 
Repenning, C. A., and R. H. Tedford. 1977. Otarioid seals of the Neogene.— Geological Survey 

Professional Paper 992, 93 pp. 
Sibley, C. G., and J. E. Ahlquist. 1972. A comparative study of the egg white proteins of non- 
passerine birds.— Peabody Museum of Natural History, Yale University Bulletin 39, vi + 276 

pp. 
Warter, S. 1976. A new Osprey from the Miocene of California (Falconiformes: Pandionidae). — 

Smithsonian Contributions to Paleobiology 27:133-139. 
Webb, S. D., B. J. MacFadden, and J. A. Baskin. 1981. Geology and paleontology of the Love Bone 

Bed from the late Miocene of Florida.— American Journal of Science 281:513-544. 
Williams, K. E., D. Nichol, and A. F. Randazzo. 1977. The geology of the western part of Alachua 

County, Florida.— Rorida Department of Natural Resources, Bureau of Geology, Report of 

Investigations No. 85, 98 pp. 



Department of Zoology, University of Florida, Gainesville, Florida 32611. 



PROC. BIOL. SOC. WASH. 

98(3), 1985, pp. 544-553 



A NEW SPECIES OF BULLFINCH (AVES: EMBERIZINAE) 
FROM A LATE QUATERNARY CAVE DEPOSIT ON 

CAYMAN BRAC, WEST INDIES 



David W. Steadman and Gary S. Morgan 



Abstract. —A new species of bullfinch, Melopyrrha latirostris, is described from 
latest Pleistocene to early Holocene cave deposits on Cayman Brae, West Indies. 
This species is larger in its cranial dimensions than other species of West Indian 
finches. Melopyrrha latirostris represents one of many species of vertebrates that 
is known on Cayman Brae only from fossils. Specimens referred to M. nigra 
taylori, which occurs today only on Grand Cayman, were recovered from the 
same stratigraphic levels as M. latirostris. 



The Cayman Islands are three small islands in the Caribbean Sea, about halfway 
between Cuba and Jamaica (Fig. 1). In 1965, Dr. Thomas H. Patton, then of the 
University of Florida, excavated several deposits of fossiliferous sediment from 

ir 

limestone caves on Cayman Brae, the easternmost island. From the richest of 
these sites, known herein as Patton's Fissure, thousands of vertebrate fossils were 
recovered from deposits of presumed Holocene age. These fossils, reported first 
by Patton (1966), are discussed in detail by Morgan (in press). 

Patton's Fissure is located in the village of Spot Bay, 3 km west of the northeast 
point of Cayman Brae at 19°45'N and 79''45'W, It is in the side of a chff about 
15 m above sea level and 250 m inland from the northern coast. Patton's Fissure 
is about 50 m long, a maximum of 4 m wide at the base, and trends east to west, 
parallel to the cliff^face. A layer of unconsolidated sediments 1-2 m deep covers 
the entire fissure. These sediments consist of buff to reddish-colored silts and 
clays, angular limestone fragments, land snail shells, and bones of small verte- 
brates. Three holes were excavated in Patton's Fissure, one of which (Hole 1) 
produced a significant amount of bone. Hole 1 was approximately 2 m square by 
1.6 m deep. The stratigraphy of Patton's Fissure is as follows: Layer 1 (0-20 cm) 
contains abundant bone, including both extinct endemic mammals as well as 
introduced species such as Rattus that indicate a post-Columbian age. Layers 2- 
4 (20-80 cm) are sparsely fossiliferous, but contained no introduced species or 
evidence of human occupation, thus indicating a pre-Columbian age for these 



and all deeper layers. Layers 5-7 (80-140 cm) are extremely rich in both land 
snail shells and bones of small vertebrates. Most of the bird fossils described in 
this paper, and the great majority of all vertebrate fossils from Cayman Brae, are 
from these three layers. Layers 8-9 (140-160 cm) contain few bones and many 
are either covered with a calcareous precipitate or are contained in an indurated 
breccia. Solid limestone was encountered below Layer 9. 

Unfortunately, neither Patton's original field notes nor the results from several 
radiocarbon ages determined in the late 1960's for Patton's Fissure are available, 
although Patton (pers. comm.) refers the age of the lower levels of this site to the 
early Holocene. This age is reasonable based upon preservation of the fossils and 



VOLUME 98, NUMBER 3 



545 



26 - 



E4" 



22"- 




L 



ORANO BAHAI4A f A 



r^EW PROVlOEf^CC 





eiEUTKERA 



ANonos 



\ ^SAN SALVADOR 

y ^CfiOOKEO 
ACKLINS '^ MAYACUANA 



ATLANTIC OCEAN 



MEXICO 




CAICOS 



COZUMEL 



18* - 



16' 




O 

L. 



100 

—I 



200 300 400 KILOMETERS 
..I t J 



LtTTL£ CAYMAN 
GRAND CAYMAN 



ST THOMAS 



CULE8RA 



PUERTO 
RICO 




,ST JOHN 
TORTOLA 

( ^A»*£GAOA 



ANGUiLLA 

^^--ST MARTIN 

* ST BARTHELEMV 



SWAH 



SABA . 



VlEQUES^C^i^ 
ST CflOM 

ST EUSTAT 
ST Kl 

NEV»S 






14-- 



^OAlteuOA 

Oantigua 

^MONTSERRAT 

^\GUA0ELOilPE 

VJ £NWftR*E CALANTE 

CSOOMINICA 

r^MARTlNIOUE 



T, LUCIA 



^t 



10 - 



ar- 




Fig. 1 . The West Indian Islands. 



upon comparisons with fossil faunas from other West Indian caves. Recently we 
dated three samples of land snails from Patton's Fissure, using the single-species 
methodology of radiocarbon age determination on land snails developed for Ja- 
maican species (Goodfriend and Hood 1983; Goodfriend and Stipp 1983). Our 
samples of Caymanian snails {Hemitrochus caymanensis) should provide a fairly 
accurate estimate of the age of Patton's Fissure, for H. caymanensis is an arboreal 
snail that does not feed on the ground (F. G. Thompson, pers. comm.). Therefore, 
this species should incorporate little if any "dead carbon" into its shell through 
ingestion of limestone. The age determinations are (in years BP, with lab number): 
11,180 ± 105 (Layer 5, SI-6 5 18); 13,230 ± 135 (Layer?, SI-65 19); and 13,850 ± 
135 (Layer 9, SI-6520). These concordant results represent maximum ages, de- 
pending upon the level at which the dated snails had incorporated environmental 
carbonate into their shells during life. The radiocarbon data suggest an age of 
latest Pleistocene or earliest Holocene for the fauna from Layers 5-9 of Patton's 
Fissure. 

The fauna from Patton's Fissure includes the extinct capromyid rodents Cap- 
romys and Geocapromys and the insectivore Nesophontes, as well as several living 
species of bats that no longer occur on Cayman Brae. Based upon Minimum 
Number of Individuals, lizards dominate the fauna of Patton's Fissure (67%), 
followed by mammals (25%, most of which are Nesophontes), and birds (8%, not 
including unidentified passerines). 




546 



PROCEEDINGS OF THE BIOLOGICAL SOCIETY OF WASHINGTON 



The avian fossils from Patton's Fissure are mainly of small passerines that 
remain incompletely studied. Conspicuous among these passerine fossils are nu- 
merous cranial elements of a finch that is much larger than Tiaris olivacea, the 
only emberizine known historically from Cayman Brae. We describe these fossils 
as representing two sympatric forms of Melopyrrha, of which one is extinct and 
the other survives only on Grand Cayman, a larger island 1 30 km west of Cayman 

Brae. 

Systematic Paleontology 

Class Aves 

Order Passeriformes 

Family Fringillidae 

Subfamily Emberizinae 

Genus Melopyrrha 

In possessing the following characters, the series of fossils from Patton's Fissure 
may be referred to Melopyrrha rather than to the closely related West Indian 
emberizine genera Tiaris, Loxipasser, Loxigilla, Euneornis, or Melanospiza. (De- 
scriptive terminology follows Baumel et al. 1979; fossil specimens are deposited 
in the Vertebrate Paleontology Collections of the Florida State Museum [UF], 
while modem skeletal specimens are from the National Museum of Natural His- 
tory, Smithsonian Institution [USNM], 

Maxilla.— In lateral aspect, more arched (curved) along both the dorsal and 
ventral surfaces (most closely approached by Loxipasser); relatively broad medial 
bar of Os nasale; presence of a small but distinct foramen in medial portion of 
the lateral bar of Os nasale, near the dorsal margin of the nares; relatively much 
shorter and stouter than in Euneornis campestris or Melanospiza richardsoni. 

Mandible.— In dorsal aspect, distal end of pars symphysialis less pointed; in 
lateral aspect, dorsal surface of dentary more curved than in all except Tiaris 
bicolor and Loxipasser anoxanthus; mandibular foramen relatively small; overall 
much stouter and more "finch-like'' than in Euneornis campestris. 

Quadrate. —Except for differences in size, it is difficult or impossible to distinguish 
individual quadrates among the six closely related genera of West Indies finches 
mentioned above. The fossil quadrates from Cayman Brae differ from those of 
the only other similarly- sized nine-primaried oscine in the fossil deposit (the 
tanager Spindalis zena; Thraupinae) in forming an obtuse angle in lateral aspect 
between the processus quadrat ojugalis and condylus squamosus, this angle being 
more nearly 90° in 

In describing Melopyrrha taylori from Grand Cayman, Hartert (1896) doubted 
the distinctness of Melopyrrha from other (unspecified) genera of finches. Standard 
check-lists (i.e., Bond 1956, Paynter 1970, AOU 1983) recognize all six of the 
emberizine genera discussed above, although we believe that most or all of these 
genera can be accommodated in an expanded genus Tiaris Swainson, 1827, on 
the basis of plumage and osteology. These finches represent an unrecognized 
evolutionary radiation within the West Indies, in many ways comparable to that 
of emberizines in the Galapagos Islands. This West Indian emberizine radiation 
and its systematic ramifications have not yet been fully documented, so we will 
describe the new species from Cayman Brae in the genus Melopyrrha rather than 
in Tiaris. 





VOLUME 98, NUMBER 3 



547 




Fig. 2. Maxillae and skulls of fossil and modem Melopyrrha, Lateral aspect in left column, dorsal 
aspect in right column. A, M nigra nigra male, USNM 321962, Cuba; B, M. n. taylori male, USNM 
554483, Grand Cayman; C, M, latirostris, holotype, UF 23011, Cayman Brae; D, M latirostris, 
paratype, UF 61022, Cayman Brae. Scale bar = 1 cm. 




548 



PROCEEDINGS OF THE BIOLOGICAL SOCIETY OF WASHINGTON 



Melopyrrha latirostris, new species 

Figs. 2, 3 

//o/o^ype. — Complete maxilla, UF 23011, from Hole 1, Layer 7 of Patton's 
Fissure, Spot Bay, Cayman Brae, Cayman Islands. Collected by T. H. Patton 
during the summer of 1965 (exact date unknown). 

Paratypes. —All from Hole 1, Layer 7 of Patton's Fissure. 12 complete or nearly 
complete quadrates, UF 6 1 008-6 1 1 9; 9 complete or partial maxillae, UF 6 1 020- 
61028; 13 incomplete mandibles, UF 23012, 61029-61042. 

Referred material, —All from Hole 1 of Patton's Fissure. Layer 2— partial max- 
illa, UF 23016. Layer 4— partial maxilla, UF 23015. Layer 5— complete or nearly 
complete quadrates, UF 61001, 61003, 61005. Layer 6— partial mandible, UF 
23013. 

Diagnosis, —Larger than Melopyrrha nigra, especially in width of maxilla and 
height of mandible (Table 1, measurements A, B, D, F, G). Nares relatively small 
compared to size of entire maxilla. In lateral aspect, ventral surface of mandible 
nearly straight (M. nigra with a distinctly obtuse angle at junction of Os suran- 
gulare and Os dentale). In lateral aspect, dorsal surface of Os dentale relatively 
straight. Os dentale proportionately long relative to length of entire mandible. 

Etymology, —From the Latin latus, broad, and rostrum, bill or snout. The name 
latirostris is regarded as a noun in apposition. 



Discussion 

Evolution, —Melopyrrha nigra, the only living species in the genus, occurs today 
on Cuba and the Isle of Pines (M. n, nigra) and Grand Cayman {M, n. taylori). 
Melopyrrha latirostris is much closer in size to M, n, taylori than to M n. nigra 
(Table 1), and on this basis it is likely that M. latirostris evolved from a population 
of M, n, taylori or its immediate progenitor that became isolated on Cayman 
Brae. Nevertheless, several of the fossils from Hole 1 of Patton's Fissure are much 
too small to be referred to M. latirostris and are similar in size to modem spec- 
imens of Af. n, taylori (Table 1). These smaller specimens, which we refer to M. 
n, taylori, include a mandible (UF 61045) from Layer 5, and a mandible (UF 
61043) and a quadrate (UF 61006) from Layer 6. From these same layers are six 
other specimens that are intermediate in size between M, latirostris and Af. n, 
taylori (a maxilla [UF 23014], a mandible [UF 61046], and two quadrates [UF 
61002, 61004] from Layer 5, and a mandible [UF 61044] and quadrate [UF 
61007] from Layer 6). We cannot say with certainty whether these last specimens 
represent very small female individuals of M, latirostris, or very large male in- 
dividuals of M, n, taylori, or hybrids between the two species. 

No specimens of M, n, taylori were recovered from Layer 7, the most fossil- 
iferous layer collected. This fact suggests that M, latirostris was already established 
on Cayman Brae before M. n. taylori colonized (or re-colonized) the island. The 
intermediate specimens suggest that genetic interchange may have occurred be- 
tween M. latirostris and M, n. taylori at that time, and the two specimens of M. 
latirostris from Layers 2 and 4 suggest that this species may have outlived its 
congener, only to disappear as well sometime in the Holocene. We regard M. 
latirostris as a full species rather than a subspecies of M. nigra because of its 



VOLUME 9 8 , NUMBER 3 



549 




Fig. 3. Mandibles and quadrates of fossil and modern Melopyrrha. Dorsal aspect of mandibles in 

I. 

left column, lateral aspect of quadrate in right column. A, M, nigra nigra male, USNM 321962, Cuba; 
B, M, n. taylori male, USNM 554483, Grand Cayman; C, M. n. taylori-M. latirostris intermediate 
fossil, UF 61046, Cayman Brae; D, M. latirostris paratype, UF 23012, Cayman Brae; E, M. latirostris 
paratype, UF 61008, Cayman Brae. Scale bar = 1 cm. 



sympatry with M. n. taylori and because of its very large size; it is larger relative 
to M. n, taylori than the latter is to M, n, nigra (see ratios in Table 1). 

The maxilla, quadrate, and mandible of M. latirostris are broader and more 
massive than in other West Indian finches. All or nearly all of the diagnostic 




550 



PROCEEDINGS OF THE BIOLOGICAL SOCIETY OF WASHINGTON 



Table 1.— Measurements (in mm) of fossil and modem Melopyrrha, giving mean, sample size (in 
parentheses), and range. UF catalogue numbers are given in parentheses for individual fossils referred 
either to M. nigra taylori or to M\ n. taylori-M. latirostris intermediates. 





A 

Maxilla: minimum 

width of medial bar 

of Os nasale 



B 

Maxilla: minimum 

width of lateral bar 

of Os nasale 



C 

Maxilla: length from 

nares to tip of Os 

premaxillare 




M. nigra nigra 
Cuba, males 

M. n. taylori 
Grand Cayman, males 

M, n. nigra 
Cuba, females 

M. n. taylori 
Grand Cayman, females 

M, latirostris 



1.3(7) 
1.1-1.7 

1.7 (7) 

1.5-1.9 

1.4(2) 
1.2-1.6 

1.5(5) 

1.4-1.7 

2.5 
2.2-2.8 




0.5 (8) 
0.4-0.7 

0.6 (8) 
0.5-0.8 

0.6 (2) 
0.6 

0.6 (5) 
0.5-0.6 

1.2(9) 
1.1-1.5 




6.1 (7) 
5.8—6.5 

7.0 (8) 
6.7-7.4 

5.8 (2) 
5.7-5.9 

6.5 (5) 
6.3-6.7 

8.2+ (3) 
8.0+-8.4 + 



M. n. taylori 
fossils, Cayman Brae 

Intermediate fossils, 
Cayman Brae 

Ratio of mean in M. n. nigra to that in 
M. n, taylori (males) 

Ratio of mean in M. n. taylori males to 
mean in M. latirostris 



2.1 (UF 23014) 0.9 (UF 23014) 7.0+ (UF 23014) 



0.76 



0.68 



0.83 



0.50 



0.87 



0.86 or less 



"+" after a value for certain fossil specimens means that the measurement of a slightly damaged 
specimen approaches to within 0.4 mm or less the actual value of the measurement if the specimen 
had been undamaged. 



characters of M. latirostris are associated allometrically v^ith its large size. The 
large, rounded maxilla of M, latirostris is reminiscent of that found in Geospiza 
crassirostris, a frugivorous emberizine finch from the Galapagos. The maxilla of 
M. latirostris, is more powerfully built than that of G. crassirostris, especially in 
the nasal region, so M. latirostris may have subsisted on a mixed diet of fruit and 
seeds. Alternatively, M. latirostris may have been mainly a seed-eater, for its 
larger bill w^ould have permitted it to take a variety of seeds. Further speculation 
on the feeding habits ofM. latirostris av^aits better documentation of the feeding 
habits of living M. nigra. The only report we have found on this topic is Johnston's 
(1975:300) for M. n. taylori of Grand Cayman, where insects and unidentified 
seeds were found in the stomachs of four different birds. 

Zoogeography, —The two species of Melopyrrha from Cayman Brae are most 
closely related to M. n. nigra from Cuba, reflecting the dominant zoogeographic 
pattern seen in the extinct and living vertebrate fauna from the island. In the 
combined vertebrate fauna from five cave deposits excavated on Cayman Brae, 
17 (81%) of the 21 species of known zoogeographic affinities are conspecific with 
or are derived from Cuban species, whereas the remaining four species have 
Jamaican affinities (Morgan, in press). Several physical and biological factors favor 
Cuba over Jamaica as a source area for most of the vertebrate fauna of Cayman 
Brae. These include: the considerably larger area, and longer coastline of Cuba; 



VOLUME 98, NUMBER 3 



551 



Table 1.— Extended* 












D 
Maxilla: 
maximum 
width 


E 

Mandible: 
total length 


F 

Mandible: 

maximum 

height of 

Os sur- 

angulare 


G 

Mandible: height at 

junction of 

Os dentale and 

Os surangulare 


H 

Mandible: maximum 

width of entire 

Os dentale 


I 

Mandible: 

length of 

cotyla 

lateralis 


J 

Quadrate: 

total 

height 


5.6 (8) 


18.2(5) 


3.9 (7) 


2.7 (8) 


8.0 (7) 


2.2 (8) 


5.2 (7) 


5.0-6.1 


17.4 18.6 


3.5 4.4 


2.3-3.1 


7.8-8.5 


2.1-2.4 


5.0-5.3 


6.3 (7) 


19.6(5) 


5.0 (8) 


3.4 (8) 


9.5 (6) 


2.6 (8) 


5.6 (8) 


6.1 6.6 


19.2-20.1 


4.5-5.3 


3.2-3.6 


9.2-9.9 


2.4-2.7 


5.5-5.9 


5.5 (2) 


17.2(2) 


3.8 (2) 


2.4 (2) 


7.4 (2) 


2.0 (2) 


5.0 (2) 


5.4-5.6 


17.1-17.3 


3.8-3.9 


2.4 


7.3-7.5 


2.0 


4.9-5.2 


5.8 (5) 


18.6(2) 


4.4 (4) 


3.1(4) 


8.5 (3) 


2.4 (5) 


5.4 (4) 


5.7-5.9 


18.3-18.9 


4.3 4.5 


2.9-3.2 


8.2-8.7 


2.2—2,6 


5.3-5.5 


8.3+ (8) 


23.4(1) 


6.5 (2) 


4.7 (8) 


11.4(8) 


3.0 (4) 


7.1 (15) 


7.6+-8.8 


23.4 


6.4-6.6 


4.1-5.0 


10.3-12.3 


2.8-3.2 


6.6-7.5 


• 






3.1 (UF 61043) 

3.2 (UF 61045) 


8.7 (UF 61045) 


1 


6.1 (UF61006) 


7.0+ (UF 23014) 







3.8 (UF 61044) 

3.9 (UF 61046) 


^^^^^ 





6.4 (UF 61002, 
61004,61007) 



0.89 



0.93 



0.78 



0.79 



0.84 



0.85 



0.93 



0.76 or less 



0.84 



0.77 



0.76 



0.83 



V' 



0.87 



0.79 



the closer proximity of Cuba to the Cayman Islands during Pleistocene glacial 
intervals; Cuba's greater species diversity; and, today's prevailing currents favor 
overwater dispersal from Cuba rather than Jamaica. Cayman Brae is almost 
equidistant (200 km) from Cuba to the east and northeast and from Jamaica to 
the southeast, and is separated from both islands by oceanic depths in excess of 
1000 m, eliminating the possibility of land bridges during the late Tertiary, How- 
ever, during periods of lower sea level in Pleistocene glacial intervals, Cuba would 
have extended to within 100 km of Cayman Brae as the extensive carbonate bank 
areas along its southern coast became exposed. The lack of evidence for a land 
connection leaves overwater dispersal as the only means by which Cayman Brae 
could have received its vertebrate fauna. Based on the low percentage of endemic 
species and the absence of generic level endemism on Cayman Brae, we believe 
that the majority of the fauna arrived during the Pleistocene. 

Extinction. —Forty species of vertebrates have been identified from Holocene 
cave deposits on Cayman Brae (8 species of reptiles, 23 of birds, and 9 of mam- 



mals), the great majority of which are from Hole 1 of Patton's Fissure (Morgan, 
in press). Of these 40 species, 17(11 species of birds and 6 of mammals) no longer 
occur on Cayman Brae. Of these 17 species, six are still found on Grand Cayman, 
seven no longer occur in the Cayman Islands but exist elsewhere in the West 
Indies, and four (including Melopyrrha latirostris) are extinct species known only 




552 PROCEEDINGS OF THE BIOLOGICAL SOCIETY OF WASHINGTON 



from the Cayman Islands. The stratigraphy and chronology of the j&ve caves 
excavated on Cayman Brae are not known well enough to determine precisely 
when the majority of these 17 species disappeared from the island. Fossils of all 
six species of mammals now extinct on Cayman Brae have been collected in caves 
from surface remains that are believed to be less than 500 years old based upon 
the presence of Rattus. On the other hand, only one of the 1 1 extirpated species 
of birds, Puffinus Iherminieri, has been recovered from these same surface layers, 
while the remaining 10 species are known only from the pre-Columbian strata in 
Hole 1 of Patton's Fissure. From the data available, we cannot determine whether 
these 1 species, which include both species of Melopyrrha, disappeared from 
natural causes before AD 1500 or were extirpated as a result of extensive habitat 
disturbance by post-Columbian peoples. 

There is no evidence of aboriginal occupation of any of the Cayman Islands 
(Hirst 1910; Richards 1955), so all Holocene habitat alteration can be attributed 
to post-Columbian settlers. Thus it is conceivable that the extinction of either or 
both forms of Melopyrrha on Cayman Brae was an historic event, but we need a 
refinement of the chronology of the upper sediments at Patton's Fissure or other 
fossil sites before the chronology of extinction of Melopyrrha on Cayman Brae 
can be resolved beyond "probably late Holocene." Nevertheless, the stratigraphic 
evidence suggests that M. latirostris may have swamped out M, n, taylori through 
interbreeding. 

That two congeneric finches could co-inhabit an island as small as Cayman 
Brae is not extraordinary, for until recently a parallel situation existed on St. Kitts 
in the Lesser Antilles. Two species of bullfinches, Lc?x/^z7/(2 noctis and L, porto- 
ricensis grandis, occurred on St. Kitts until several decades ago when L, p. grandis 
apparently became extinct (Olson 1984). St. Kitts is not much larger than Cayman 
Brae in area, but is much higher in elevation, supporting lush forest in the volcanic 
highlands. However, Olson points out that L. p. grandis may have evolved in the 
lowlands of St. KLitts, so habitat diversity may have played little if any role in 
permitting the two species of Loxigilla to co-exist on St. Kitts. Although the 
limestone forest of Cayman Brae has a low canopy height and a low species 
diversity today, birds elsewhere in the West Indies tend to be relatively abundant 
in both absolute numbers and numbers of species in arid habitats (Kepler and 
Kepler 1970; Pregill and Olson 198 1). Thus we see no reason why the prehistoric 
forests of Cayman Brae could not have supported two or more species of con- 
generic finches. 

Acknowledgments 

We profited from discussions with G. A. Goodfriend, S. L. Olson, and F. G. 
Thompson. R. Stuckenrath provided radiocarbon ages. We thank V. E. Krantz 
for the photographs, and M. E. Parrish for their help in labelling. E. M. Paige 
drew Fig. 1 . Comments by R. I. Crombie, H. F. James, and S. L. Olson improved 



the manuscript. 



Literature Cited 



American Ornithologists' Union. 1983. Check-list of North American birds, 6th ed. Allen Press, 

Inc., Lawrence, Kansas, 877 pp. 
Baumel, J. J., A. S. King, A. M. Lucas, J. E. Breazile, and H. E. Evans (eds.). 1979. Nomina Anatomica 

Avium. Academic Press, London, 664 pp. 



VOLUME 98, NUMBER 3 



553 



Bond, J. 1956. Check-list of birds of the West Indies. Academy of Natural Sciences of Philadelphia, 

214 pp. 

Goodfriend, G. A., and D. G. Hood. 1983. Carbon isotope analysis of land snail shells: implications 

for carbon sources and radiocarbon dating.— Radiocarbon 25:810-830. 
, and J. J. Stipp. 1983. Limestone and the problem of radiocarbon dating of land-snail shell 



carbonate.— Geology 11:575-577. 
Hartert, E. 1896. Description of a new finch from the West Indies.— No vitates Zoologicae 3:257. 
Hirst, G. S. S. 1910. Notes on the history of the Cayman Islands. P. A. Benjamin Manufacturing 

Company, Kingston, Jamaica, 412 pp. 
Johnston, D. W. 1975. Ecological analysis of the Cayman Island avifauna. — Bulletin of the Florida 

State Museum, Biological Sciences 19(5):235-300. 
Kepler, C. B., and A. K. Kepler. 1970, Preliminary comparison of bird species diversity and density 

in Luquillo and Guanica forests. Pp. E- 183-1 86 in H. T. Odum, ed., A tropical rain forest: a 

study of irradiation and ecology at El Verde, Puerto Rico. U.S. Atomic Energy Commission, 

Division of Technical Information, Oak Ridge, Tennessee. 
Morgan, G. S. [In press.] Late Pleistocene and Holocene fossil vertebrates from the Cayman Islands, 

West Indies.— Bulletin of the Rorida State Museum, Biological Sciences. 
Olson, S, L. 1984. The last St. Kitts bullfinch {Loxigilla portoricensis grandis) and the extinction of 

its race.— Bulletin of the British Ornithologists' Club 104(4): 12 1-1 23. 
Patton, T. H. 1966. Occurrence of fossil vertebrates on Cayman Brae, B. W. I.— Caribbean Journal 

of Science 6(3-4): 181. 
Paynter, R. A., Jr. 1970. Subfamily Emberizinae. Pp. 3-214 in R. A. Paynter, Jr., ed.. Check-list of 

birds of the world, vol. 13. Museum of Comparative Zoology, Harvard University, Cambridge, 

Massachusetts. 
Pregill, G. K., and S. L. Olson. 1981. Zoogeography of West Indian vertebrates in relation to 

Pleistocene climatic cycles.— Annual Review of Ecology and Systematics 1981, 12:75-98. 
Richards, H. G. 1955. The geological history of the Cayman Islands.— Notulae Naturae 285:1-11. 






(DWS) Department of Vertebrate Zoology (Birds), Smithsonian Institution, 
Washington, D.C. 20560. Present address: New York State Museum, 3140 Cul- 
tural Education Center, Albany, New York 12230; (GSM) Florida State Museum, 
University of Florida, Gainesville, Rorida 32611.